From Precision to Scale: Operationalizing Nuanced Care Models in a Rapidly Evolving Health Ecosystem with Charlie Harp

·23 min·2 clips

Charlie Harp warns that AI on bad data is 'artificial stupidity,' a risk that could lead to serious patient harm.

The episode is an interview with Charlie Harp, CEO and founder of Clinical Architecture, conducted by host Saul at the HLTH event. Harp begins by reflecting on his 35-year career and the growing industry focus on data quality. He explains that Clinical Architecture was started 18 years ago to address the foundational 'plumbing' of healthcare data, necessary for interoperability and meaningful use cases like AI, clinical decision support, and value-based care. Harp emphasizes that without good data, even the best technologies fail, coining the phrase 'artificial intelligence on bad data is artificial stupidity.' He discusses the Patient Information Quality Improvement Framework (PICI), an open-source project developed with partners like Levitt Partners, the VA, and CMS, designed to objectively measure data quality and build trust in interoperability. Harp argues that data quality is a 'lifestyle choice' for the industry, not a one-time project. The host asks about various stakeholders: payers need seamless integration of claims and clinical data, health systems require data governance and liquidity (exemplified by a client's ability to locate ventilators in 30 seconds during a hurricane), and life sciences face normalization challenges for research. Harp contrasts healthcare with finance, noting that finance demands penny-perfect accuracy while healthcare has historically tolerated critical data errors. He concludes that healthcare is too complex and 'disrupt-proof' for quick flips, requiring evolutionary steps rooted in data quality. The episode ends with Harp promoting his company's website and his podcast, InfraMonster.

As heard by us

A payer-data conversation that treats quality as infrastructure, not cleanup.

Charlie Harp, CEO and founder of Clinical Architecture, anchors the discussion in a plain but durable idea: AI is only as useful as the data behind it.

Read the full review in PlayNext →

Why you'd press play

You get a blunt case for cleaner healthcare data before AI makes bad data worse.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from The Beat